Value Proposition With No Mandate
You are a business architect embedded in a 4,000-person logistics firm. You have built a credible capability model and a value-stream view of order-to-delivery. The COO, your sponsor, is enthusiastic but has no budget authority over the two divisions that would need to change. The divisional MDs are measured on cost-per-parcel and on-time delivery, and see your work as overhead. One MD has offered a small pilot: 'map our depot onboarding process, and if it's useful we'll talk'. You have 0.5 FTE of analyst support and a quarterly window before budget round. The enterprise architecture team already has a process inventory that overlaps yours and is protective of it.
1. How should you choose between pushing the enterprise-wide model and accepting the narrow pilot?
Read the worked answer
Take the pilot - but choose its scope strategically rather than accepting the offered scope passively.
The enterprise model has no path to a decision: the sponsor cannot fund it, and the two people who can are not persuaded. Pushing it consumes your quarter and produces another document. The pilot is the only route to a funded decision, because it converts an MD from skeptic to reference customer - and in a cost-per-parcel culture, a peer's result moves budget faster than a COO's endorsement.
But 'depot onboarding' as offered is likely too narrow to demonstrate anything an MD cares about. Negotiate the scope up one notch: map depot onboarding as part of order-to-delivery, so the pilot shows the handoff and the downstream effect on on-time delivery - the MD's own metric. That keeps the pilot small enough to deliver in the window while making the output speak the MD's language.
Explicitly agree, up front, what 'useful' means and what happens if it is: a named follow-on decision and a date. Otherwise 'we'll talk' is the whole contract.
2. How do you handle the overlapping process inventory owned by enterprise architecture?
Read the worked answer
Do not compete with it and do not silently duplicate it. Duplication is the fastest way to make your work look like overhead, which is precisely the accusation you are trying to defeat.
Practical steps:
- Inventory the overlap explicitly - list which of your elements already exist in their inventory and which are genuinely new (typically: value streams, capability-to-process linkage, and the cross-divisional handoffs, which process inventories usually do not capture).
- Lead with the complement, not the replacement. Position your work as adding the why and the cross-boundary view to their what. Most EA process inventories are strong on cataloguing and weak on end-to-end value flow - that is your opening.
- Propose a single source of truth with split ownership: they own the process catalogue, you own the value-stream and capability layers, with a maintained mapping. Get this agreed in writing before the pilot output is published, not after - after publication it becomes a territorial negotiation.
- If they refuse integration, reference their inventory by identifier in your pilot output rather than restating it. It costs you little and makes refusal to collaborate visible without you having to say it.
The pilot's credibility depends on it looking like the firm's work, not a parallel universe. Reuse is cheaper than authorship here, and it buys an internal ally you will need at budget round.